GLACIERS OF UNAKWTK INLET AND COLLEGE FIORD 299 It is impossible to say how much Harvard Glacier may have retreated from 1794 to 1887 and 1898, for the maps are not detailed enough. Applegate's map, made in 1887, seems to indicate that the terminus of Harvard Glacier had about the same relationship to the cascading glaciers on the western side of College Fiord as when Gannett made his more accurate map in 1899. The glacier seems to have changed very little between 1899 and 1905. An advance took place, however, between Grant's first visit in 1905 and his second visit four years later,1 estimated to have been half a mile on the western side of the glacier and quarter of a mile on the eastern side. Gannett's map, made in 1899, when compared with the Higgins map ten years later, shows clearly that the advance could not have been so great as this without great retreat from 1899 to 1905. In 1910 the junior author compared photographs taken in 1899, 1905, and 1909 with, conditions in the field and is inclined to think there was little if any retreat from 1899 to 1905, and to place the advance between 1905 and 1909 at a much smaller figure than Grant does, probably not more than two hundred yards. Between 1909 and 1910 the continuation of this advance amounted to only 100 or 150 feet. This recent advance is shown clearly in photographs, which bring out the forward movement of the east side, and the overriding of a narrow barren zone on the western edge. The inclined moraines of the terminus were also pushed farther over by an advance of Radcliffe Glacier at the same time. The advance in 1910 resulted in increased crevassing in the lateral moraines on either margin of Harvard Glacier near the terminus, and in the overriding and destruction of forest on each side. This advance was actually in progress when we visited Harvard Glacier on July 17 to 22, 1910. One of the pronounced effects of the advance was the increase in wave work as the result of more active discharge of icebergs. By these icebergs waves a delta, built by the stream from Downer Glacier, a quarter of a mile from the glacier front was being cut away at a very rapid rate. A precipitous wave cut cliff in the gravels of this delta had a height of 40 feet in 1910, most of the cutting having taken place during the previous year. At the western edge of Harvard Glacier (PL CXV), where the advance in 1910 seemed to be due to activity of Radcliffe tributary, a push moraine was being formed on the beach. Along the glacier margin, a short distance to the northward, the moraine was made up largely of an inextricable tangle of macerated fragments of trees and roots, mixed with soil, moss, peat, gravel, and till. There were well-developed peat rolls, and in places the push moraine was 15 feet high. One of the trees which has just been overturned was a spruce 12 inches in diameter and probably over 100 years old, in-, dicating that the glacier had not previously advanced as far as the present stage for at least a century. On this side of the fiord as on the eastern side there has been an increase in activity of iceberg waves which has resulted in the cutting of little cliffs, and the uprooting of shrubs. Cascading Glaciers of College Fiord—General Description. On the western side of College Fiord there are seven cascading glaciers, as well as a number of much smaller ice masses. These Mendenhall2 described in 1898 as follows: "Numerous small glaciers, easily distinguished from the unconsolidated snow by the blue color of their fronts and crevasses, hung from the summits, or extended down the slopes." i Giant, U. S. and Biggins, D. F., Bull. Amer. Geog. Soc., Vol. "XT-Trr, 1911, p. 825. « Mendenhall, W. C., Twentieth Annual Beport, U. S. Geol. Survey, Part VII, 1900, p. 272.